4.7 million people in the United States had Alzheimer's disease in 2010, a number that is projected to nearly triple by 2050. While the National Plan to Address Alzheimer's Disease and Related Dementias is an effort to coordinate federal, state, and local efforts to confront this growing challenge, it explicitly states ?this is a national plan and not a federal plan?, with individual states left to determine how to best care for these patients and their caregivers. Because of cognitive and functional changes as age-related dementia progresses, the needs of these patients and their caregivers extend far beyond healthcare, requiring a diverse response from states. However, the care most accessible to community-dwelling patients with age-related dementia is direct medical care. In addition, as a patient's ability to direct their own care declines, the healthcare provided to them may be driven by the needs and preferences of the caregiver. These patients then experience: fragmented medical care, poorly-coordinated across multiple outpatient providers; potentially preventable hospitalization; and overuse of antipsychotics despite extensive evidence of harms. Because of the state-led nature of the National Plan, it is critical to provide states with key predictors of this inappropriate care to help guide their policy. We will use national Medicare data, a national survey of older adults with dementia and their caregivers, and an Expert Panel of researchers and state policy experts to complete the following aims: (1) Identify patient and community factors associated with potentially inappropriate care delivered to community- dwelling adults with age-related dementia and establish accurate national and state-level estimates of this care. We will use a cross-sectional Medicare 20% sample to determine use of the following for all 50 states: a) potentially-preventable hospitalization; b) fragmented outpatient care; and c) antipsychotic use. (2) Determine the contribution of additional patient clinical, functional, caregiver, and caregiving characteristics to potentially inappropriate care relative to the effect of location. We will use the National Health and Aging Trends Study, National Survey of Caregivers, and Area Health Resource File to determine the patient, caregiver, and regional healthcare system characteristics associated with potentially inappropriate medical care for persons with dementia. (3) Develop an evidence-based policy making guide for dementia that we will use to interview state aging policy officials. An Expert Panel of dementia researchers and state aging policy experts will apply our Aim 1 and Aim 2 findings to develop an evidence-based policymaking framework with particular emphasis on populations most at risk, which we will use to guide interviews with state aging policy officials. The impact of our work will be to: 1) identify the patient, caregiver, and regional factors that contribute to potentially inappropriate medical care for patients with age-related dementia; 2) provide a baseline against which to measure future system-level interventions; and 3) develop a framework for states to develop evidence-based dementia care state plans.